Circular elements or opposed points

You have to use opposed points for local size on width features (because there are no circular elements - widths are planar).  For cylinders, opposed points will simulate a calliper check which means it will not necessarily capture the form error.  Circular elements simulates a ring/plug gauge.  This is the section from the GDT training...

How would circular elements simulate a plug gauge for an inner diameter?  The slide uses an inner diameter as an example and says that it is the smallest circumscribed diameter, but that doesn't make sense to me for an inner diameter.  This is why I don't see any use for circular elements.  Unless the slide has the two mixed up and it's meant to be the largest inscribed diameter for ID and the smallest circumscribed diameter for OD.  That would make sense to me, but that's not how the slide has it.

For an internal hole, the UAME is simulating a plug gage being put into the hole - it is the external to material size (inscribed diameter in this case).  Local size with circular elements is the smallest circumscribed diameter that would fully encompass the points - essentially the internal to material size.  Yes, I know that you wouldn't be able to physically fit a ring gage over those points but circular elements can be useful in detecting when there's a lot of form error that 2-point local size could miss.

So conversely then, would circular elements be the smallest inscribed diameter for an OD?

Yes

Ok, one more question then.  Would opposed points be the largest distance between two points for an ID and the smallest distance between two points for an OD?  Thanks for the help.

Yes

Thank you.  

I love this type of discussion!  Great question E. Neil, thanks for your wealth of knowledge! Sometimes pcdmis training tells you what box or options to click bit you don't always get the magic behind the why's. This forum is a gold mine! 

I just found this slide from another presentation that illustrates how circular elements captures the form error of a tri-lobed diameter whilst opposed points does not...

Thank you Neil.  That is very helpful. But as far as I can see, the size dimension would only report the 15.970 with circular elements.  How would I get the 16.030?

Thank you Neil.  That is very helpful. But as far as I can see, the size dimension would only report the 15.970 with circular elements.  How would I get the 16.030?

It would depend on whether it was an internal (hole) or external (stud) diameter.  If it was a hole, the UAME would be 15.970 and the circular element local size would be 16.030.  If it was a stud, the UAME would be 16.030 and the circular element local size would be 15.970.

If this was a stud, then wouldn't the UAME take the form error of the entire length of the stud into account as well, whereas the local size would only be a circular cross section?  Please forgive my crude drawing here, but if the stud was wavy, then in theory the local size could measure ok in every cross section, but the UAME would be much larger.  I guess what I'm trying to ask is how would you know from the UAME whether the form error is the result of a local size being too large or if the stud had a wavy shape to it?  Or is that another topic entirely?  

ESchertz said:

I guess what I'm trying to ask is how would you know from the UAME whether the form error is the result of a local size being too large or if the stud had a wavy shape to it?

You wouldn't - that would need to be determined by adding a cylindricity and looking at the graphical analysis.  Looking at the difference between UAME and local size just gives you an indication of the AMOUNT of form error present  It doesn't give any indication of the shape of the feature, it's just telling you what the maximum and minimum extremes of size are.

Thank you.  I guess I wish there was an option to show both the largest inscribed and the smallest circumscribed local size with circular elements along with the UAME.  Or maybe I'm just not understanding it correctly.  Local size with circular elements is a 2D circular cross section, isn't it?